>And the IQ regresses back to the mean of the parents, so no chance of the kid being a genius. I could have misunderstood this, but I thought regression toward the mean meant *on average* the offspring tend to have IQs closer to the mean of the parents' race than the parents. So even two genius white people who have a kid are likely to have an average IQ kid. The reason anti-racemixers cite regression towards the mean is to show that, for example, a very smart black person (100 IQ) and a normal white person (100 IQ) are more likely to have offpsring with IQs around 90 even though the black parent is smarter than average. But the normal bell curve applies. They could still have a kid that's smarter or dumber than that. >Polar bears and grizzly bears produce fertile offspring. They are different species. When you mix grizzly and polar bears, you're getting rid of the unique adaptations of each bear. As a counterexample, one wolf specie is capable of interbreeding with coyotes, and have started producing coywolves in the northeastern US. While these offspring *should* lose the special adaptations of both parent species, they also gain adaptive advantage due to a mixing of the parent traits (namely being larger size than coyotes and more comfortable living in food-rich, manmade environments than wolves).